A major challenge in the field of aging and dementia is the inability to predict asymptomatic individuals who will develop dementia and, in particular, Alzheimer's disease. Abnormal brain changes precede clinical cognitive changes by 10 to 20 years providing a window of opportunity to prevent dementia. The National Institute on Aging and the Alzheimer's Association published guidelines to characterize the pre-clinical and mild cognitive impairment (MCI) stages of Alzheimer's disease, placing a major emphasis on biomarkers. However, we have limited information on the prevalence of biomarkers at the population-level in mid-life and late- life, and on their ability to predict cognitive outcomes. In addition, the assessment of abnormal brain changes in asymptomatic individuals using imaging biomarkers may not be a cost-effective screening approach for secondary prevention and may not be applicable to all persons. The few risk scores that have been proposed thus far to predict dementia do not assess the risk of MCI and do not consider biomarker abnormalities. The broad, long-term goal of this renewal application is to develop tools to predict and prevent cognitive decline and dementia. To accomplish this goal we propose 4 specific aims: 1) To estimate the prevalence of neuroimaging biomarkers such as brain amyloid accumulation, neurodegenerative pathology, and vascular pathology in a defined population; 2) To prospectively examine the association of these imaging biomarkers with cognitive outcomes; 3) To develop risk scores to predict biomarkers and cognitive outcomes; and 4) To provide data and materials for related projects. To successfully address these gaps in knowledge and to accomplish our study goal, this project will capitalize on 2 unique and established resources available in the Olmsted County, MN population. The population-based and prospective Mayo Clinic Study of Aging has enrolled nearly 4,000 participants to date; of whom over 1,000 have extensive neuroimaging biomarkers of amyloid deposition and neurodegeneration. The Rochester Epidemiology Project medical records-linkage system is used to recruit population-based samples, to identify demographics and clinical risk factors, to study cognitive outcomes for subjects who could not be studied face-to-face, and to compute weights to adjust for non-participation. At the completion of this project, we will have estimated the population-based prevalence of neuroimaging biomarkers, determined their predictive value for cognitive outcomes, developed a risk score to predict imaging biomarkers, and determined the added predictive value of a risk score that includes neuroimaging biomarkers. These findings will provide critical information for public health planning, endpoints for therapeutic trials, and cost effective tools for the early detection of brain abnormalities and cognitive outcomes.